276 Dr. C. R. A. Wright on the Determination of 



day, when a number of careful observations of the current 

 passing and the value of E were made at as early an hour as 

 possible (usually about 9 o'clock), so that the effect of the heat- 

 convection currents due to alteration of laboratory tempera- 

 ture should be a minimum. The particular temperature of 

 the measuring resistance-coil (platinum-silver or German- 

 silver wire) being known, its resistance in ohms R x was known; 

 and the potential-difference in volts E / being known by stan- 

 dardizing the scale of the electrometer by a Clark's cell, the 

 current was known in C.G.S. units, being given by the formula 



E' x 10 8 

 C =^5 ry-g. The resistance of the voltameter, R, being known, 



xx-Y X 1U 

 the value of e = E — CR was thus known (usually the term 

 CR was negligible, or at least small compared with E). The cur- 

 rent was then allowed to pass for 24 hours, and the observation 

 repeated as early as possible in the day, and so on. Usually 

 the results of consecutive days' measurements agreed extremely 

 closely. When a sufficient number of daily determinations had 

 been made, the resistance in circuit was altered so that a dif- 

 ferent current passed, and another series of determinations 

 lasting over some days made, and so on. Sometimes after 

 making various series of determinations of the relative values 

 of C and e in this way, the value of C was brought back to 

 nearly the same value as that possessed in observations of 

 several days or weeks before. In such cases the values de- 

 duced in the second set of observations were sensibly the same 

 as those formerly obtained ; or at least they uniformly corre- 

 sponded to points lying sensibly on the same mean curve, 

 obtained by plotting the currents as abscissae and the values of 

 e as ordinates. Thus, for instance, the average of three closely 

 accordant sets of consecutive morning readings with volta- 

 meter no. 1 (containing acid of which 100 cub. centims. con- 

 tained 40*0 grammes H 2 S0 4 ), made near the beginning of a 

 long series of determinations lasting over many weeks, and 

 that of three other sets of accordant observations made near 

 the end of the period, gave the following numbers — 



Earlier series. 



Later series. 



Current in micro- 

 webers. 



e in volts. 



Current in micro- 

 webers. 



e in volts. 



23-5 

 210 

 20-6 



1-906 



1-892 

 1-880 



22-3 

 20-1 

 198 



1-901 

 1-895 

 1-870 



Average 21 7 



1893 



20-7 



1-889 



thus giving the final average — Current = 21'2 ; e=l , 891. 



